Deployment blueprint
A city's waste programme fails quietly: the contracts team, the operations desk, and the BI analyst each answer the same question differently. We put bins, vehicles, crews, contractors, and lots on one data model, so the answer stops depending on who you ask.
Built for this scale
The challenges
Three pressures arrive together — operational, contractual, and environmental — and the data needed to answer any of them sits in a different system.
Challenge
Hundreds of vehicles, tens of thousands of bins, and several contractors produce data the authority cannot consolidate. Operations quotes one collection figure, contracts quotes another, and the monthly review spends its first hour reconciling spreadsheets instead of deciding anything.
Challenge
If the only record of a completed round is the contractor's own logbook, enforcement is voluntary. Disputed line items accumulate to every invoice cycle, and the authority pays for work it has no way to verify.
Challenge
Policy commits the city to measurable cuts in fleet emissions. But kilometres driven, litres burned, and tonnes lifted are captured loosely or not at all, so the target lives in a document rather than in the field.
Challenge
Bins are emptied on a fixed calendar, so one district overflows while a half-loaded truck drives past full containers in the next. The signal that triggers the next collection is a citizen phoning in, which means the city is always one step behind.
How it would work
We instrument first and model second. Fill, weight, and motion sensors go onto collection points; telematics goes onto the fleet — including vehicles the authority does not own. Every one of those feeds lands in the same event pipeline, so a contractor's truck and a municipal truck are the same kind of object by the time they reach the console. On top of that sits the operating model: bins, vehicles, crews, lots, contracts, and departments as related records rather than separate systems, with the city map and the org chart expressed as the same hierarchy.
The architectural point
The design decision that matters is that we treat the programme as brownfield from day one. Existing hardware is adopted, not replaced, and new sensor classes, vehicle types, and analytics attach to a live programme without a migration project. That is what lets the same console later absorb camera-based street inspection or wet-weather dispatch on the fleet that is already connected — the expensive part, getting every asset onto one model, is already done.
Trucks live
312
Bins > 80%
1,480
Off-route
7
Lots on plan
94%
City waste · live
liveDistrict 4 · overflow cluster
18 bins above threshold
Contractor B · route deviation
Veh. 2214 · 2.6 km off lot
Night shift · plan accepted
412 collection points
Tonnes lifted · last 12 days
Illustrative console view — sample data, not a live operation.
Platform capabilities
Composed from modules that already run in the platform — geospatial, real-time, compliance, analytics, and workforce on one console.
Every vehicle, bin, and shift on one continuously updating map. Geofence and route-adherence rules raise an exception the moment it happens, not in the overnight batch.
ExploreUltrasonic sensors report container levels around the clock, so the day's route is built from which bins are actually full rather than which day of the week it is.
ExploreEach contractor's committed plan is scored against telemetry daily, with field evidence attached to the line item. The invoice meeting starts from a shared record.
Utilisation, vehicle age, capacity, compaction ratio, and expected volume combine into how many trucks a district needs — a number the authority can defend in a contract negotiation.
The day's optimised plan reaches the cab with navigation, task list, and proof-of-collection capture. Supervisors run inspections from the same app against the same records.
ExploreThe existing route network imports as KML and becomes live, monitored geometry — districts, lots, and corridors you can alert on and report against.
ExploreHardware in the field
Bin sensors, vehicle telematics, in-cab video, and edge inference, all normalised into one event stream before they reach the console.
Retrofitted into existing containers with no civil works. Continuous telemetry is what converts a fixed schedule into collection-by-need.
ExploreGPS, engine data, and fuel signals across the collection fleet, including contractor vehicles. Device-agnostic — registered by IMEI, paired to a SIM, assigned to a vehicle.
ExploreIn-cab and body-mounted video gives route-adherence and compliance events a picture attached, so a disputed exception is settled by looking at it.
ExploreNavigation, task tracking, and proof-of-collection in the cab, with a supervisor mode for inspection rounds on the same hardware.
On-vehicle inference flags bin overflow, dumped construction debris, and unpermitted signage during a normal round — the collection fleet doubles as an inspection fleet.
Integrations
The platform sits on top of the systems the authority already runs, connecting at the contract, finance, and identity boundaries.
Vehicles already running another vendor's box are ingested rather than re-fitted. The authority gets one picture across mixed fleets and keeps its supplier options open.
Compliance scores flow to the finance system so every invoice line carries the field evidence behind it, and credit notes stop being a negotiation.
Console access maps to the authority's directory over SAML, so department structure and staff turnover are handled where they already are.
Camera detections are checked server-side against the permit registry, so a flagged hoarding or skip arrives with its licence status already resolved.
FAQ
Short answers here — the detailed version is a conversation whenever you're ready.
Discuss your deployment
Sovereign-data-friendly. Enterprise multi-org architecture. Built to integrate with the devices, SIM providers, and back-office systems you already run.
